0
$\begingroup$

For a Lambertian surface, the radiance at a point is

$$L = \frac{\alpha}{\pi}\cos\theta_i E_0$$

If we have an infinite flat surface, albedo = 1, illuminated directly normal, then we get

$$L = \frac{E_0}{\pi}$$

But now suppose we are sitting very close to the surface. The surface occupies $2\pi$ sr of our view, so the irradiance we measure is

$$E_r = 2\pi L = 2E_0$$

Is this correct? I was expecting to get $E_0$ back again. $2E_0$ makes it feel like we're violating the conservation of energy or something.

$\endgroup$
1
$\begingroup$

I don't really understand what are you doing. I think your first equation shouldn't have a cos factor in it.

We have the relation,

$BRDF = dL_r / dE_i$

That is the brdf is the ratio of reflected radiance to incoming "irradiance". Re-arranging this gives us,

$ dL_r = BRDF * dE_i$

For diffuse surfaces we know,

$BRDF = \alpha/\pi$

Substituting in above equation we have,

$dL_r = \alpha/\pi * dE_i$

Irrespective of what angle we view, the radiance is the same. Now for a diffuse surface with albedo = 1 we have what you said,

$dL_r = dE_i/\pi$

Now in order to measure Irradiance, we can use the relation,

$E = \int_{2\pi} L \cos(\theta) d\omega$

$E = \displaystyle\int_{\phi = 0}^{2\pi} \displaystyle\int_{\theta = 0}^{\pi/2} L \cos(\theta) \sin(\theta) d\theta d\phi$

$E = E/\pi * \pi$

$E = E$

So you got back the same irradiance.

| improve this answer | |
$\endgroup$
  • $\begingroup$ The first cosine is for the direction of incoming light, to get the "true" irradiance or whatever. E.g. $E_0$ coming in at a $90^\circ$ angle doesn't light up the surface at all. I think my discrepancy comes from the fact you integrated $L\cos(\theta)$ instead of just integrating $L$. $\endgroup$ – HiddenBabel May 23 '19 at 3:59
  • $\begingroup$ 1). I know. But if you are putting in the cosine then you have to do it properly. $dE$ will change into $dL_f$. Because as shown below $dE = dL \cos(\theta) d\omega$ 2) Why would you just integrate $L$. You get irradiance by integrating $L \cos(\theta)$. $\endgroup$ – gallickgunner May 23 '19 at 6:34
  • $\begingroup$ Ah I thought I had already taken care of the $\cos\theta$ when I was calculating solid angle for my actual problem (not a flat plane). But that's just it, I needed $\cos^2\theta$ in total, and now I get $E=E$. $\endgroup$ – HiddenBabel May 23 '19 at 8:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.